Remission of the polycystic ovarian condition (PCO) in the rat following hemiovariectomy.код для вставкиСкачать
THE ANATOMICAL RECORD 226:328-336 (1990) Remission of the Polycystic Ovarian Condition (PCO) in the Rat Following Hemiovariectomy MARGARET CONVERY, GERALD F. McCARTHY, AND JAMES R. BRAWER Department of Anatomy and The McGill Centre for the Study of Reproduction, McGill University, Montreal, Canada ABSTRACT The estradiol valerate-induced polycystic ovarian condition in the rat represents a normal ovarian response to aberrant endocrine stimuli. Although we have shown that removal of one polycystic ovary (hemiovariectomy) results in restoration of cyclicity and normal morphology in the remaining ovary by 1 week, nothing is known about the process of recovery or about the role of the hypothalamopituitary unit in initiating recovery. We have therefore examined ovaries at 3, 12, 24,48, and 320 hours following removal of the contralateral polycystic ovaries. The ovarian content and size distribution of healthy and atretic follicles was determined, a s well a s the occurrence of follicular cysts, type I11 large follicular structures, and corpora lutea. The plasma LH pattern was also examined a t a short postoperative interval. At 3 hours, there was a significant increase in mean ovarian weight t h a t coincided with the emergence of healthy large secondary follicles. By 12 hours, there was a significant sustained diminution in the number of atretic follicles of all sizes, but the total number of healthy follicles did not increase significantly until 120 hours. The cystic follicles had all but disappeared by 120 hours because of mechanical compression by newly developing ovarian tissue. Ovarian recovery is, therefore, biphasic, consisting of a very early diminution in atresia coincident with, and perhaps caused by, a major alteration in the plasma LH pattern. The second phase is characterized by a wave of follicular recruitment and development. Anovulatory acyclicity characterized by ovaries containing multiple cystic follicles is a reproductive anomaly that occurs in a wide range of mammalian species under a variety of circumstances. It is a major form of infertility in the human, and it plagues the pork, cattle, and sheep industries. The polycystic ovarian condition (PCO) can be produced in the laboratory rat by a variety of experimental manipulations, including constant light exposure (Daane and Parlow, 19711, anterior hypothalamic deafferentation (Halasz, 1969; Blake et al., 1973), and neonatal androgen treatment (Gorski, 1971). We have generated a chronic PCO condition in the rat by giving a single large dose of estradiol valerate (EV) (Brawer et al., 1978; Hemmings e t al., 1983; Schulster e t al., 19841, and we are currently using this model to explore the pathogenesis and the biology of the polycystic ovary. The ovaries in the EV-treated rat contain no corpora lutea and few healthy secondary follicles. Large follicular cysts occur, which are characterized by a highly attenuated membrana granulosa and a hypertrophied thecal cell layer comprised of large polygonal, lipidfilled cells. Clusters of large polygonal secondary interstitial cells are abundant in the stroma (Hemmings et al., 1983; Schulster et al., 1984; Brawer et al., 1986, 1989). It appears that the expression of this characteristic polycystic morphology does not reflect intrinsic ovarian pathology, but rather is the response of a n essentially normal ovary to a n abnormal pattern of gonado0 1990 WILEY-LISS, INC tropin stimulation (Brawer et al., 1986; McCarthy et al., 1986; Grosser e t al., 1987; Carriere e t al., 1989). Examination of the histological transition from the normal to the polycystic state (Brawer et al., 1986, 1989) indicates that the polycystic morphology is really the product of complete or arrested atresia of developing secondary follicles and the absence of ovulation (Brawer et al., 1986). The loss of a n ovulatory LH surge accounts for the lack of corpora lutea, and the atresia of secondary follicles results in the numerous clusters of hypertrophied lipid-containing secondary interstitial cells. The cysts themselves represent partial or arrested atresia of large secondary follicles or type I11 large follicular structures (Brawer et al., 1986, 1989; Desjardins and Brawer, 1989). The processes of follicular recruitment and atresia that produce the polycystic ovary seems to be driven by a specific plasma gonadotropin pattern (McCarthy et al., 1986, 1987; Grosser et al., 1987). We have shown that cyclicity and normal ovarian morphology can be restored in EV-induced PCO by the removal of one polycystic ovary (i.e., hemiovariectomy) (Farookhi et al., 1985), indicating that the polycystic ovary retains the capacity for normal function. Regular Received April 6, 1989; accepted June 20, 1989. Address reprint requests to James R. Brawer, Department of Anatomy, 3640 University Street, Montreal P.Q. H3A 2B2, Canada. REMISSION OF PCO 329 vaginal cyclicity and normal ovarian histology are re- Montreal, Quebec, Canada) and then housed one per stored by 1 week following hemiovariectomy. The pres- cage. Atrial catheters were flushed with 50 IU of hepence of corpora lutea at this time point denotes the arin sodium (Heparin Sodium Injection U.S.P., Allen & restoration of a gonadotropin surge. Interestingly, this Hanburys, Montreal, Quebec, Canada) diluted in 0.5 recovery process occurs in the absence of any signifi- ml of bacteriostatic (1.5%benzyl alcohol) 0.9%saline cant change in the mean plasma gonadotropin concen- (Bacteriostatic Sodium Chloride Injection U.S.P., trations (Farookhi e t al., 1985). We have hypothesized, Squibb Canada) every 2-3 days thereafter. After a postoperative interval of 1 week, three anitherefore, that hemiovariectomy interrupts the PCOassociated plasma LH pattern (McCarthy et al., 1986, mals were each placed a t 0400 into a n isolation cage. A 1987; Grosser et al., 19871, resulting in a new pattern 64 cm length of PE-100 tubing (Clay Adams, Parsippany, NJ; 0.034 inches in I.D. and 0.060 inches in O.D.) that supports complete follicular development. At present, however, we have no information on the was fed into the cage through the bore of a tightly process of ovarian recovery from the polycystic state. coiled spring and attached to the catheter. The spring Since we know that it is essentially complete by 1 week was held in place by a pivot a t the top of the cage that following hemiovariectomy (Farookhi e t al., 19851, it is afforded the animal free movement during the samclear that during this short interval major changes in pling procedure. At this time, 1hour prior to sampling, follicular dynamics occur resulting in development of the tubing and catheter (sampling line) were flushed follicles to the preovulatory state as well as in disap- with 50 IU of heparin sodium diluted in 1 ml of bactepearance of the cysts. To determine how and why this riostatic 0.9%saline. Beginning a t 0500 hours, 0.5 ml blood samples were occurs, we have examined the changes in follicle populations at early intervals following hemiovariectomy drawn every 10 minutes over a 4 hour period. Each in animals with PCO. In view of the significance of the blood sample was centrifuged for 2 minutes in a Beckplasma LH patterns in PCO (McCarthy et al., 1986, man Microfuge (15,600g). Two 100 p,1 plasma aliquots 1987; Grosser et al., 1987), we would predict that any were removed and frozen on dry ice. The remaining change in follicular populations should be preceded by blood cells were resuspended in 200 p1 of bacteriostatic a change in the plasma LH pattern. We have, there- 0.9% saline containing 10 IU heparidml. The resusfore, also examined the plasma LH pattern several pended blood cells were injected into the animal hours after hemiovariectomy. Since we have found no through the sampling line a t the end of the next sam*correlation between the FSH pattern and the develop- pling interval. The samples from each animal were ment and expression of the polycystic morphology (Mc- stored at -80°C for later determination of LH. Carthy et al., 1986, 1987), we have not included obserSurgical Procedures vations on the FSH pattern. At 0900 hours the animals were removed from the MATERIALS AND METHODS isolation cages, anesthetized with ether, and the right Plasma LH Patterns ovary of each animal was removed through a flank incision. At 1200 hours the animals were returned to Animals and treatment the sampling cages and were prepared for serial blood Twelve young, female Wistar rats (145-200 g) were sampling. At exactly 1300 hours (4 hours posthemiobtained from Charles River Ltd. (St. Constant, Que- ovariectomy) 0.5 ml blood samples were withdrawn at bec, Canada). They were housed in groups of four and 10 minutes intervals over a 4 hour period a s described were given free access to pelleted rat food and water. above. The animals were exposed to 14 hours of light daily The remaining (unsampled) three animals were (lights on a t 0700 hours), and their estrous cycles were sham operated a t 0900 hours to serve as controls for the monitored by daily examination of vaginal smears. hemiovariectomized group. At 1000 hours each animal Only those animals that displayed two consecutive nor- was placed into a n isolation cage and prepared for semal 4 day estrous cycles prior to treatment were used rial blood sampling. One hour later (2 hours after sham in the study. operation) the animals were sampled as described Each animal was injected (i.m.) with a 2 mg dose of above. The samples from each animal were stored a t estradiol valerate (EV) (Delestrogen U.S.P., Squibb -80°C for later LH determinations. Canada Inc., Montreal, Quebec, Canada) dissolved in 0.2 ml sesame oil. Three to 4 weeks thereafter, all an- LH assay imals displayed persistent vaginal cornification. Only The concentrations of LH were determined by radiothose animals that continued to show unequivocal and immunoassay using 100 pl samples. Samples (pre- and sustained vaginal cornification by week 7 after injec- posthemiovariectomy and of the sham-operated anition were selected for the study. mals) were assayed in singleton for LH in separate assays. Assay kits were obtained from the National Catheter placement and serial blood sampling Pituitary Agency (NIAMDD, Bethesda, MD). The The six animals that had been selected a t week 7 NIAMDD reference preparation used for the assays were each fitted with a chronic indwelling atrial cath- was LH-RP-2. The assay procedures have been deeter using a modification (Grosser et al., 1987) of a scribed previously (Grosser et al., 1987). The intraastechnique previously described by Tannenbaum and say coefficient of variation in the LH assay for the preMartin (1976). Following the surgery, each animal was and posthemiovariectomy samples a t the 20%, 50%, given a n intramuscular injection of 0.35 ml of 300,000 and 80%level of counts for a n arbitrary dose relative to IU/ml penicillin G procaine (Ayercillin, Sterile Penicil- that of a zero dose (BIBo) was 4.9%,3.1%, and 10.4%, lin G Procaine Suspension U.S.P., Ayerst Laboratories, respectively. The limit of detection for LH was 105 pg/ 330 M. CONVERY ET AL. ml. The intraassay coefficient of variation for the LH assay used for samples from sham-operated animals was 8.8%, 6.3%, and 26.3%, respectively, at the 20%, 50%, and 80% level of counts (B/Bo). The limit of detection of LH in this assay was 70 pg/ml. Data analysis LH pulses were defined, for the most part, using the criteria established by Gallo (1981) and later modified by Ellis and Desjardins (1982). In our pulse analysis, a n LH determination and associated coefficient of variation (CV) was assigned only to the lowest-most point (nadir) of the ascending phase of a potential pulse. An LH value was also assigned to the highest-most point (peak) of a potential pulse. A pulse was defined when the peak value was greater than the nadir value plus two times the CV of the nadir value. The lowest-most point of the descending limb of a pulse, as long as it differed from the peak value plus twice its associated CV, marked the end of a pulse. Pulse nadir, peaks, amplitudes, and frequencies were each determined for the 4 hour sampling period and were expressed as the mean ? the standard error of the mean. Morphology Treatment of animals Young female Wistar rate, weighing between 145 and 200 g, were purchased from Charles River Ltd. The animals were maintained under conditions of lighting described above. Estrous cycles were monitored for 2 weeks by daily examination of vaginal smears. Animals each exhibiting at least two normal 4 day cycles were used in this study. Each animal was lightly anesthetized with ether and given a n intramuscular injection of 2.0 mg EV dissolved in 0.2 ml sesame oil. Eight weeks after the EV treatment, a group of 30 animals each demonstrating unequivocal and sustained cornified vaginal smears during the 2 prior weeks were chosen for the study. Animals were anesthetized with ether, and the right ovary from each animal was removed through a flank incision. The ovaries were examined to confirm that the animals used in the study did, in fact, have PCO a t the time of ovariectomy. Several such ovaries contained corpora lutea and hence disqualified the animals from the study, leaving only four (out of the six) animals in one of the groups. In order that the sample number be consistent, we have used only four animals, with confirmed PCO, from each postoperative time point. Groups of four animals were killed by decapitation at 3,12,24,48, and 120 hours after hemiovariectomy. The remaining ovary was removed, weighed, and fixed in Bouin’s solution for 24 hours. The oviducts were removed and examined for the presence of ova. Following fixation, the ovaries were dehydrated and embedded in paraffin. Each ovary was serially sectioned a t 7 pm, and the sections were stained with hematoxylin and eosin (Brawer et al., 1986). Four polycystic ovaries removed a t the time of hemiovariectomy were prepared as described above. Analysis of follicle populations The procedure for the analysis of follicle populations used in this study is based on a previously published methodology (Brawer et al., 1986). All primary and secondary follicles in each ovary were measured using a n ocular micrometer. Measurements were made in the section of the follicle that contained the nucleus and nucleolus of the ovum. For each follicle, two perpendicular diameters were measured. Each diameter was measured from basement membrane to basement membrane, and the average of these two measurements was then calculated. In addition to size designation, each follicle was classified as either healthy or atretic. A follicle was considered atretic if it exhibited degeneration of the ovum or one or more pyknotic granulosa cells. Follicle size distribution histograms were generated for each postoperative time point. The follicles at each time point were grouped into three size categories for the purpose of statistical analysis. These are follicles with diameters less than 250 pm (mostly primary and very early secondary follicles), follicles with diameters between 250 and 500 pm (secondary follicles), and follicles with diameters greater than 500 pm (large secondary follicles, including putative preovulatory follicles). Multigroup comparisons of the data were analyzed by one way analysis of variance. Comparison of means between two groups was evaluated by using Student’s t test. Differences were considered significant at P < 0.05. The larger follicular structures such as cystic follicles were identified and counted a t each posthemiovariectomy time point. Cystic follicles were identified according to established criteria (Brawer et al., 1986, 1989; Desjardins and Brawer, 1989). Type I11 large follicular structures (Brawer et al., 1989; Desjardins and Brawer, 1989)were distinguished from large secondary follicles by virtue of the fact that they did not contain ova, although occasionally a very degenerate circular structure, possibly a deteriorated ovum, occurred floating free in the large antrum. The membrana granulosa was thick and often plicated. Also, in contrast to large secondary follicles, large polygonal cells were frequently seen interspersed among the smaller fusiform variety in the theca interna. RESULTS Plasma LH Patterns The LH patterns for the nonoperated controls are shown in Figure 1A-C. The patterns in these animals are identical to the PCO-associated LH pattern described previously (Grosser et al., 1987). The mean nadir, peak, and amplitude of LH pulses in the nonoperated controls were 123 S 18, 204 ? 35, and 78 2 45 pg/ml, respectively. The frequency was approximately one pulsehour. The plasma LH patterns displayed by the three sham-operated controls (not illustrated) indicate that the procedure itself may cause a significant suppression of LH pulsatility. Two of the three shamoperated controls showed no detectable LH pulses over the sampling interval, while the third animal exhibited three pulses of amplitude and duration comparable to those characterizing pulses in nonoperated controls. Within the 4 to 8 hour interval following hemiovariectomy, the plasma patterns of LH differed markedly from those observed prior to hemiovariectomy (Fig. 1D-F). The mean nadir, peak, and amplitude of pulses in the hemiovariectized animals were 248 L 93,406 2 331 REMISSION O F PCO 6ool 400 A 6oo] B 6o01 D ,:I OJ, 6ool 16000 . * , , 1 I 80001 01 I 0 , i . , . . . . : . . j O , F 1 400 2o-l , 0 24000 r I 80 160 240 0 80 160 240 Fig. 1. Plasma LH patterns. A-C: LH patterns prior to hemiovariectomy in three individual animals with PCO. D-F: LH patterns in the same animals 4 hours after hemiovariectomy. The scale in F differs from the others to accommodate a very large LH episode. As a result, the smaller pulses cannot be resolved in this panel. All apparent pulses meeting the criteria described in Materials and Methods are indicated by a n asterisk. 150, and 194 k 130 pg/ml, respectively. The pulse frequency was similar to that observed in the nonoperated control patterns (about one pulsehour), as were the durations and shapes of the pulses. 3.68 0.45, 3.33 0.54, and 5.02 0.48 g, respectively. In each group, the ovaries removed after hemiovariectomy were significantly heavier than their respective controls (removed at hemiovariectomy) (P < 0.05). Cyclicity All animals at the time of hemiovariectomy exhibited persistent vaginal cornification. This condition was also observed in all animals at the 3,12,24, and 48 hour postoperative intervals. By 120 hours, there was considerable variation in the smear patterns. The smears of two animals exhibited all cell types, one exhibited a n estrous-diestrous smear, and one exhibited a proestrous-estrous smear. No ova were observed in the oviducts at any postoperative time point. Ovarian Weights The weights of the ovaries removed at the various time points were compared with those removed at hemiovariectomy (controls). The mean weight of control (polycystic) ovaries ( C SEMI removed at hemiovariec0.370 g. The mean ovarian weight tomy was 1.57 3 hours following hemiovariectomy (2.505 0.540 g) was significantly greater than the control weight (P < 0.05). The mean ovarian weights a t 12,24,48, and 120 hours posthemiovariectomy were 3.37 k 0.70, * * * * Ovarian Content of Corpora Lutea, Cysts, and Type 111 Large Follicular Structures Corpora lutea were absent from control ovaries and from ovaries removed at 3 and 12 hours following hemiovariectomy. At 24 hours, one or two corpora lutea appeared in three (out of four) ovaries, and a t 48 hours two of the ovaries showed one or two corpora lutea. Three of the ovaries removed at 120 hours each exhibited four to six large corpora lutea, whereas one ovary contained only one corpus luteum. Smaller clusters of luteinized tissue, possibly resulting from luteinization of small secondary follicles also occurred in each ovary (see Fig. 6). Type I11 large follicular structures (Figs. 2,3A) were easily identified on the basis of previously established criteria (Brawer et al., 1989; Desjardins and Brawer, 1989). The use of serial sections in the present study allowed us to confirm our suspicion that type I11 large follicular structures do not contain ova. An average of 332 M. CONVERY ET AL. Fig. 2. Type 111 large follicular structure. The large antrum is surrounded by a thick membrana granulosa, which is plicated along the right side of the field. x 100. six type I11 large follicular structures occurred in each control ovary removed a t hemiovariectomy. This number declined to a n average of two per ovary a t 12, 24, and 48 hours. By 120 hours posthemiovariectomy, only one of four ovaries showed a single type I11 large follicular structure. Seven to 10 cystic follicles (Figs. 3B, 4) occurred in each ovary removed a t 3, 12, and 24 hours after hemiovariectomy. At the 48 hour interval, each ovary contained between four and seven cysts, and a t 120 hours cysts were absent from all but one ovary, which contained only two cystic follicles. The cysts a t the later time points appeared squeezed and distorted by developing ovarian tissue (Fig. 5). This process of compression was progressive in time, ultimately resulting in the collapse and disappearance of most cysts by 120 hours (Fig. 6). Ovarian Follicle Content The mean number of healthy and atretic follicles (Fig. 3C) in control (polycystic) ovaries (time 0) and in ovaries removed at 3 hours posthemiovariectomy were similar (Fin. 7). At 12 hours there was a reduction in the number Of follicles largely because Of a decrease in the population of atretic There was little change a t 24 and 48 hours, but at 120 hours the Fig. 3.The wall of A: a type 111 large follicular structure, B: a cyst, and C: an atretic secondary follicle. In each case, the antral surface of the membrana granulosa is indicated by arrowheads. x 400. Fig. 4. Polycystic ovary charaterized by numerous cystic follicles and by a n absence of corpora lutea and of healthy secondary follicles. The hilus appears in the upper right of the field. X 17. Fig. 6. Ovary removed 120 hours after hemiovariectomy exhibiting the full complement of follicles that typifies the normal ovary. Several corpora lutea are also evident. x 17. 0 Fig. 5. Ovary removed 48 hours after removal of contralateral ovary (hemiovariectomy).Two residual cysts (stars) are compressed and distorted by developing follicles and only one cyst (triangle) maintains the characteristic appearance. The hilus appears in the right of the field. ~ 1 7 . 3 12 24 TIME (hrs) 48 120 Fig. 7. Total ovarian content of follicles a t the different posthemiovariectomy intervals. The mean ( 5 SEMI number of healthy follicles (of all sizes) a t each posthemiovariectomy time point is depicted by the stippled bars. The mean ( 2 SEM) number of atretic follicles are represented by the striated bars at the top, Size Distribution of Follicles mean number of follicles exceeded that seen in controls. The mean number of atretic follicles/ovary remained relatively constant from 12 to 120 hours, but there was a marked increase in the population of healthy follicles. The mean follicular content of the ovaries included follicles in all phases of development from the early primary (small) to the late secondary (large) stage. The data have been further broken down to analyze the follicular size distribution at each postoperative inter- 334 M. CONVERY ET AL. "1 40 A 48 hrs 0 hrs 1-/ 40 1 20 0 U LL LI 60 B W 1 2 hrs 80 D 1 2 0 hrs m 52 60 40 40 20 20 0 , 0 . 100 50 ' 150 200 . 250 . 300 . 350 . 400 . . 500 450 0 200 100 ) 50 150 300 250 500 400 350 450 FOLLICLE DIAMETER (,urn) @ Fig. 8. Size distribution of healthy and atretic follicles at four posthemiovariectomy time points. The mean ( 2 SEMI number of follicles in the size ranges indicated on the X axes are depicted. Healthy follicles are indicated by the stippled bars. and the atretic follicles are designated by the cross-hatched bars a t the top. val. This permitted the evaluation of the contribution of the different follicular populations to the total mean number at each time point. Size distributions of follicles with diameters less than 500 pm are presented in Figure 8. Figure 8A shows the size distribution of follicles in the polycystic (control) ovaries. The pattern at 3 hours posthemiovariectomy (data not shown) was essentially identical. At 12 hours (Fig. 8B), however, there were significant deviations from the control pattern. There was a significant decrease in the number of atretic follicles of all sizes (P < .001), and this low level of atresia was maintained throughout the duration of the study. Although there were fewer primary follicles (diameter <250 pm) at 12 hours than in controls, the difference was not significant. However, a significant increase in this population occurred between 12 (Fig. 8B) and 48 (Fig. 8C) hours, and this trend continued such that the number of these small follicles (diameter <250 pm) was significantly greater at 120 hours (Fig. 8D) than in controls. No significant changes were observed in follicles with diameters of 250-500 (small-medium secondary follicles) at any of the postoperative time points. The control ovaries contained no healthy large sec- 181 0 3 12 24 TIME (hrs) 48 120 Fig. 9. Ovarian content of large secondary follicles a t the different posthemiovariectomy intervals. The mean ( 5 SEM) numbers of healthy large secondary follicles (diameter > 500 km) a t each time point is depicted by the stippled bars. The mean (? SEM) numbers of atretic large secondary follicles are designated by the striated bars. REMISSION OF PCO ondary follicles (Fig. 9). By 3 hours, however, each ovary contained a n average of six such structures. Another significant increase in this population occurred between 3 and 24 hours (P < 0.001), after which no further change occurred. There were no significant differences in the number of atretic large secondary follicles between any of the time points. 335 sizes between 0 and 12 hours is a key event in the process of recovery. Atresia plays a major role in the development of the polycystic condition (Brawer et al., 1986). The process of atresia is responsible for the low overall content of follicles in the polycystic ovary, as well a s the dearth of healthy large secondary follicles. Moreover, it appears that the cysts themselves are products of arrested atresia (Brawer et al., 1986). It is, DISCUSSION therefore, not surprising that the recovery from the Within several hours following hemiovariectomy, polycystic condition is preceded by a reversal of the marked changes occur in both the plasma LH pattern trend favoring atresia to one permitting the normal and in the histology of the polycystic ovary. The LH full range of follicular development. patterns were determined from 4 to 8 hours following Although the polycystic morphology appears to rethe procedure, leaving open the question as to whether sult from a predominance of atresia, the attenuation of alterations in the LH pattern may actually occur a t a n atretogenesis is insufficient, by itself, to account for even earlier time point. This is, however, a likely pos- full ovarian recovery. There is a significant hiatus besibility, since significant changes in ovarian weight tween the reduction of atresia (at 12 hours) and the and histology are apparent a s early a s 3 hours posthe- increase in the total number of healthy follicles obmiovariectomy. Hemiovariectomy produced marked served at 120 hours. Indeed, the ovarian content of increases in nadir, peak, and amplitude of LH pulses healthy follicles a t 12 hours is less than in polycystic despite the fact that anesthesia andlor surgical stress controls because of reduction in the number of healthy alone resulted in a suppression of LH pulses. Clearly, primary and small secondary follicles a t 12 hours. The remove1 of one polycystic ovary caused abrupt and pro- relatively stable ovarian follicular content observed benounced feedback responses at the hypothalamopitu- tween 12 and 48 hours can be attributed to a lag beitary level. Whether this rapidly induced modulation of tween the diminution in atresia and a n increment in gonadotropin release is the principle causal antecedent follicular recruitment. By 120 hours both of these pheto the process of ovarian recovery remains to be deter- nomena contribute to the enhanced ovarian content of mined. It does, however, seem probable in view of pre- follicles. vious evidence linking a specific LH (but not FSH) patThus at least two processes contribute to the recovtern to the onset and maintenance of PCO in the EV ery of the polycystic ovary. The LH pattern is probably treated rat (McCarthy et al., 1986, 1987). Moreover, the principle determinant for the level of atresia. A alteration of the PCO-associated LH pattern by means second event, however, must account for the accelerof treatment with the opiatergic antagonist naltrexone ated folliculogenesis. This may, in fact, be a further results in significant recovery of the polycystic ovaries modification of the LH pattern, or it may involve a late within 3 days (Carrier et al., 1989). posthemiovariectomy change in the FSH pattern. AlCoincident with the rapid change in the plasma LH ternatively, the stimulus may be intraovarian, derivpattern are significant changes in the ovaries. Ovaries ing from the new population of large healthy secondary removed at the 3 hour posthemiovariectomy time point follicles. The nature of the folliculogenic stimulus reweighed significantly more than the polycystic con- mains to be determined. trols. Moreover, they contained a n abundance of The disappearance of cystic follicles appeared to be healthy large secondary follicles, which were totally caused by progressive mechanical compression resultabsent in control ovaries. Since neither the content nor ing from the development of, first, large secondary folthe size distribution of the other follicular structures licles and later, follicles of all sizes. Follicular cysts, changed from the 0-3 hour time period, the increase in therefore, do not appear to respond actively, a s do other ovarian weight is most probably due to the emergence populations of follicles or follicular structures, to hemiof the healthy large secondary follicles. ovariectomy, but rather are eliminated by passive conIf, as suggested, alterations in the plasma LH pat- striction. This is consistent with our earlier observatern contribute significantly to the changes in ovarian tions that cyst development coincides with a loss of histology, the rather sudden appearance of healthy receptivity to LH (Brawer et al., 1989), suggesting that large secondary follicles may be causally linked to the the cystic follicle is a relatively inert structure. rapid onset of the relatively large amplitude LH pulses One of the most intriguing follicular structures following hemiovariectomy. We have shown that the unique to the polycystic ovary is the type I11 large foldevelopment of the polycystic ovarian morphology is licular structure. These are comprised of a membrana preceded by a loss of high amplitude LH pulses occur- granulosa and thecal cell layers similar to those charring between 16 and 21 days following EV treatment acterizing Graafian follicles. The type I11 large follicu(McCarthy e t al., 1989). The resultant plasma LH pat- lar structure, however, differs from a Graafian follicle tern, which is maintained from 21 days onward, is that in that it is often larger, exhibits mitotic figures in the typifying the established PCO condition (Grosser et. perimural granulosa cells (Brawer et al., 1989; Desjaral., 1987). Interestingly, the incidence of atresia, partic- dins and Brawer, 1989), and, a s demonstrated by the ularly among large secondary follicles, is maximal at present study, contains no ovum. Paradoxically, i t is this post-treatment interval (Brawer et al., 1986). One the only follicular structure in the polycystic ovary in might predict, therefore, that the reintroduction of which granulosa cells bind LH (hCG) (Brawer et al., large amplitude LH pulses would rescue a cohort of 1989), and it is, therefore, the only structure that could developing follicles from atresia. be readily luteinized following a n LH surge. Two pieces The marked reduction in atresia in follicles of all of evidence suggest that the corpora lutea, which begin 336 M. CONVERY ET AL. to appear a t 24 hours following ovariectomy, derive from these structures. First, the incidence of corpora lutea increases as that of type I11 large follicular structures declines. Second, ova were never found in the oviducts, indicating that the structures giving rise to the corpora lutea did not, or could not, ovulate. The nature, function, and role of the type I11 large follicular structure in PCO remain to be elucidated. In summary, the poverty of follicles in general, and of healthy large secondary follicles in particular, in the polycystic ovary appears to be the result of a constant rate of follicular recruitment together with a high level of atresia. Hemiovariectomy reverses both of these trends and produces a rapid decline in atresia, followed somewhat later by enhanced follicluar recruitment. The resultant newly developing ovarian tissue compresses and eventually destroys the cysts. The diminution in atresia that initiates the recovery process is coincident with the restoration of high amplitude LH pulses. ACKNOWLEDGMENTS The authors gratefully acknowledge the technical assistance of Ms. Dalia Chen. This work was supported by a n operating grant to J.R.B. from the Medical Research Council of Canada. LITERATURE CITED Blake, C.A., R.J. Scaramuzzi, J . Hilliard, C.H. Sawyer 1973 Circulating levels of pituitary gonadotropins and ovarian steroids in rats after hypothalamic deafferentation. Neuroendocrinology, 1 2 3 6 97. Brawer, J.R., M. Munoz, and R. Farookhi 1986 Development of the polycystic ovarian condition in the estradiol valerate-treated rat. Biol. Reprod., 35:647-655. Brawer, J.R., F. Naftolin, J . Martin, and C. Sonnenschein 1978 Effects of a single injection of estradiol valerate on the hypothalamic arcuate nucleus and on the reproductive function in the female rat. Endocrinology, 103501-512. Brawer, J.R., M. Richard, R. Farookhi 1989 Pattern of human chori- onic gonadotropin binding in the polycystic ovary. Am. J. Obstet. Gynecol., 161:474-480. Carriere P., R. Farookhi, and J.R. Brawer 1989 The role of aberrant hypothalamic opiatergic function in generating polycystic ovaries in the rat. Can. J. Physiol. Pharmacol., (in press). Daane, T.A., and A.F. Parlow 1971 Serum FSH and LH in constant light-induced estrus: Short-term and long-term studies. Endocrinology, 88:964-968. Desjardins, G.C., and J.R. Brawer 1989 The development and maintenance of a polycystic condition in ovaries transplanted under the kidney capsule. Anat. Rec. (in press). Ellis, G.B., and C. Desjardins 1982 Male rats secrete luteinizing hormone and testosterone episodically. Endocrinology, 110: 1618-1627. Farookhi, R., R. Hemmings, and J.R. Brawer 1985 Unilateral ovariectomy restores ovulatory cyclicity in rats with a polycystic ovarian condition. Biol. Reprod., 32t530-540. Gallo, R.V. 1981 Pulsatile LH release during periods of low level LH secretion in the rat estrous cycle. Biol. Reprod., 24:771-777. Gorski, R.A. 1971 Gonadal hormones and the perinatal development of neuroendocrine function. In: Frontiers in Neuroendocrinology. L. Martini, W.F. Ganong, eds. Oxford University Press, New York, pp. 237-290. Grosser, P.M., G.F. McCarthy, B. Robaire, R. Farookhi, and J.R. Brawer 1987 Plasma patterns of LH, FSH, and prolactin in rats with a polycystic ovarian condition induced by oestradiol valerate. J . Endocrinol., 114:33-39. Halasz, B. 1969 The endocrine effects of isolation of the hypothalamus from the rest of the brain. In: Frontiers in Neuroendocrinology. L. Martini, W.F. Ganong, eds. Oxford University Press, New York. pp. 307-342. Hemmings, R., R. Farookhi, and J.R. Brawer 1983 Pituitary and ovarian responses to luteinizing hormone releasing hormone in a rat with polycystic ovaries. Biol. Reprod., 29:239-248. McCarthy, F.G., J.R. Brawer, and R. Farookhi 1986 Plasma LH patterns characterizing the development of the polycystic ovarian condition (PCO) in the estradiol valerate (EVMreated rat. Biol. Reprod., [Suppl. 11 34:153 (Abstract). McCarthy, G.F., R. Farookhi, and J.R. Brawer 1987 Polycystic ovaries occur in response to a specific pulsatile pattern of luteinizing hormone (LH). Biol. Reprod., [Suppl. 11 36:15 (Abstract). Schulster, A., R. Farookhi, and J.R. Brawer 1984 Polycystic ovarian condition in estradiol valerate-treated rats: Spontaneous changes in characteristic endocrine features. Biol. Reprod., 31 587-593. Tannenbaum, G., and J.B. Martin 1976 Evidence for an endogenous ultradian rhythm governing growth hormone secretion in the rat. Endocrinology, 12562-570.